Gene 436 (2009) 56–65

Contents lists available at ScienceDirect

Gene

j ourna l homepage: www.e lsev ie r.com/ locate /gene

A novel Drosophila antisense scaRNA with a predicted guide function

Giuseppe Tortoriello, Maria Carmela Accardo, Filippo Scialò, Alberto Angrisani, Mimmo Turano, Maria Furia ⁎Department of Structural and Functional Biology, University of Naples “Federico II”, Complesso Universitario Monte Santangelo, via Cinthia, 80126 Napoli, Italy

Abbreviations: ncRNA, non coding RNA; snoRNA, sbody; scaRNA, small Cajal bodies RNA; CAB box, Cajal bosnoRNP, small nucleolar ribonucleoprotein; dUHG, Drosogas5, growth arrest-specific transcript 5; TMG, 2,2,7-digoxigenin; JNK pathway, Jun N-terminal Kinase pathw⁎ Corresponding author. Tel.: +39 081679072 (office

679076 (lab); fax: +39 081/679233.E-mail address: mfuria@unina.it (M. Furia).

0378-1119/$ – see front matter © 2009 Elsevier B.V. Aldoi:10.1016/j.gene.2009.02.005

a b s t r a c t

a r t i c l e i n f o

Article history:

A significant portion of euk Received 12 November 2008Received in revised form 3 February 2009Accepted 3 February 2009Available online 20 February 2009

Received by C. Feschotte

Keywords:DrosophilascaRNAsnoRNACajal bodysnRNA U1bGAS5

aryotic small ncRNA transcriptome is composed by small nucleolar RNAs. Fromarchaeal to mammalian cells, these molecules act as guides in the site-specific pseudouridylation ormethylation of target RNAs. We used a bioinformatics search program to detect Drosophila putativeorthologues of U79, one out of ten snoRNAs produced by GAS5, a human ncRNA involved in apoptosis,susceptibility to cancer and autoimmune diseases. This search led to the definition of a list of U79-related flysnoRNAswhose genomic organization, evolution and expression strategy are discussed here.We report that anintriguing novel specimen, named Dm46E3, is transcribed as a longer, unspliced precursor from the reversestrand of eiger, afly regulatory gene that plays a key role in cell differentiation, apoptosis and immune response.Expression of Dm46E3 was found significantly up-regulated in a mutant strain in which eiger transcription isgreatly reduced, suggesting that these two sense–antisense genes may be mutually regulated. Relevant to itsfunction, Dm46E3 concentrated specifically in the Cajal bodies, followed a dynamic spatial expression profileduring embryogenesis and displayed a degenerate antisense element that enables it to target U1b, adevelopmentally regulated isoform of the U1 spliceosomal snRNA that is particularly abundant in embryos.

© 2009 Elsevier B.V. All rights reserved.

1. Introduction

In eukaryotic cells,many regulatorycircuitries implicated in importantaspects of development and cell differentiation involve the expression ofsmall ncRNAs (reviewed by Mattick, 2004; Mattick and Makunin, 2006).Indeed, a significant portion of the eukaryotic small ncRNA transcriptomeis composed by small nucleolar RNAs (snoRNAs). These molecules,generally ranging from 60 to 300 nucleotides in length, are present fromArchaeal to mammalian cells. With a few exceptions, snoRNAs areclassified into two major families, named box C/D and H/ACA, on thebasis of common sequencemotifs, structural features and common sets ofassociatedproteins. A fewmembers of each family are required for properrRNA processing, whereas the majority directs, by base-pairing guidingmechanisms, the two most common types of nucleotide modificationspresent on eukaryotic RNAs (reviewed by Bachellerie et al., 2002; Kiss,2002). H/ACA snoRNAs guide in fact RNA pseudouridylation, while C/DsnoRNAs direct 2′-O-ribosemethylation of target RNAs (Bachellerie et al.,2002; Kiss, 2002).Methylation guide snoRNAs display a simple structure,characterized by the presence of consensus C (5′-RUGAUGA-3′) and D(5′-CUGA-3′) motifs close to the 5′ and 3′ termini of the molecule,

mall nucleolar RNA; CB, Cajaldy-specific localization signal;phila homolog of human UHG;trimethylguanosine cap; DIG,ay.), +39 081/679071, +39 081

l rights reserved.

respectively. Additional and often degenerated internal copies of C and Delements (designated C′ and D′) may also be present and may confer tothe snoRNA the ability to guide methylation at two sites. In fact, D/D′upstream regions act as antisense elements able to select the residue tomodify through the formation of specific duplexes with the RNA target(Bachellerie et al., 2002; Kiss, 2002). According to the so called D/D′+5rule, the methylated site is invariably complementary to the snoRNAnucleotide mapping five positions upstream of the D/D′ box. In themethylation process, the methyltransferase catalytic activity is furnishedby fibrillarin, one of the four evolutively conserved core proteins thatassociate to snoRNAs of this family to compose the C/D functional smallnucleolar ribonucleoprotein (snoRNPs) complexes (reviewed by Henraset al., 2004; Kiss et al., 2006). Initially, the role of snoRNAswas thought tobe restricted to rRNA modification in ribosome biogenesis, but it is nowevident that thesemolecules can targetotherRNAs, including snRNAs andmRNAs (Henras et al., 2004). Indeed, a subgroup of snoRNAs involved insnRNA modification do not localize in the nucleolus but in the Cajalbodies, nucleoplasmic organelles inwhich concentratemany ribonucleo-protein (RNP) factors whose major function is in maturation, assemblyand/or trafficking of snRNPs, snoRNPs and transcription complexes(reviewed by Stanek et al., 2008; Pontes and Pikaard, 2008). Cajal body-specific RNAs (scaRNAs) can be of either the C/D or H/ACA type or cancomprise a H/ACA domain embedded in a C/D box structure, and aretypically involved in themethylation and/or pseudouridylation of the PolII-transcribed snRNAs of the U1 spliceosome (Richard et al., 2003; Henraset al., 2004).

During the last few years, the repertoire of snoRNA's biological rolewas continuously widening. For instance, human C/D box snoRNAs

57G. Tortoriello et al. / Gene 436 (2009) 56–65

specifically expressed in the brain have been implicated in theinterference with A-to-I editing (Vitali et al., 2005) and in theregulation of alternative mRNA splicing of the serotonin receptor(Kishore and Stamm, 2006). The human non-coding growth arrest-specific transcript 5 (GAS5), which encodes ten different snoRNAs inits introns (U44, U47, U74, U75, U76, U77, U78, U79, U80, U81; Smithand Steitz, 1998), was shown to play a critical role in the control of cellgrowth and apoptosis, to be down regulated in breast cancer(Mourtada-Maarabouni et al., 2009), and has evoked as a candidategene in the development of autoimmune disease (Mourtada-Maara-bouni et al., 2008). Together with the recent demonstration of tumorsuppressor characteristics in the human snoRNA U50 (Dong et al.,2008), these observations suggest that snoRNA genes may be involvedin controlling oncogenesis and sensitivity to therapy in cancer. Inaddition, an increasing number of orphan snoRNAs which lackantisense to rRNAs or snRNAs have been experimentally identifiedfrom different organisms. Since some of them display a developmen-tally or tissue specific pattern, and in mammals also imprinting, it hasbeen suggested that they play still uncharacterized regulatory roles ongene expression (reviewed by Rogelj, 2006).

Genomic organization of C/D snoRNA genes has been found to becomplex and variegated. In plant, yeast and protozoan the largemajority of these genes is independently transcribed, whereas innematodes monocistronic snoRNAs are almost as abundant as thoseintron-encoded (Huang et al., 2007). In Drosophila, organization andexpression of C/D snoRNA genes have been systematically investi-gated. The two genome-wide computational approaches performedso far (Accardo et al., 2004; Huang et al., 2005) were essentially basedon the primary structural elements of snoRNAs and the sequencecomplementarity to modified sites present on rRNA. However, thesescreens had to rely on the conservation of rRNA methylation sitesdescribed in other eukaryotes, since direct information on methylatedresidues present on the fruit fly rRNA is still missing. Nonetheless, theywere effective and largely contributed to furnish an overall view oforganization of snoRNA genes in the Drosophila genome. Similarly tovertebrates, most of fly C/D snoRNAs are intron-encoded and usuallyarranged in the mode of one snoRNA per intron (Yuan et al., 2003;Accardo et al., 2004; Huang et al., 2005), whereas only a very few areindependently transcribed monocistrons. Altogether, bioinformaticsand molecular approaches have so far led to annotation of 63 fly C/Dbox snoRNAs, which are predicted to account for about two-thirds ofthe rRNA methylations and a very few snRNAs modifications,suggesting that a substantial number of Drosophila snoRNA genesstill await to be recognized.

In a previous analysis (Accardo et al., 2004), we used the SnoScanprogram (Lowe and Eddy, 1999) and sequences of yeast rRNAmethylation sites conserved in Drosophila to search for methylationguide snoRNAs in the fruit fly genome. To extend further thisapproach, we performed a new screen using the same program butsequences of human rRNA methylation sites conserved in Drosophila.Among these conserved sites, we specifically selected for the analysisthose that were methylated on human, but not on yeast rRNA. As afirst outcome, we report here the results of our quest of U79-relatedfly snoRNAs. Human U79 is responsible for methylation of 28S at a sitehighly conserved in metazoans, but unmodified in yeast, andrepresents one out of ten snoRNAs encoded by the GAS5 pro-apoptoticncRNA involved in cell growth, cancer and autoimmune diseases(Smith and Steitz,1998; Nakamura et al., 2008;Mourtada-Maarabouniet al., 2008, 2009). This search led to the identification of two novelsnoRNAs with peculiar genomic arrangement, one of which wasantisense to eiger (egr), a Drosophila regulatory gene encoding aTumor Necrosis Factor-like ligand involved in the regulation of celldifferentiation, death and immunity (Igaki et al., 2002; Kauppila et al.,2003; Schneider et al., 2007). Besides adding further information onthe repertoire, evolution and expression strategy of Drosophila C/DsnoRNAs, the detailed characterisation of this antisense snoRNA

provides a glimpse into the molecular mechanisms underlyingsnoRNA biological functions, as well as those governing egr regulation.

2. Materials and methods

2.1. Fly strains

Canton S was used as wild-type strain in all experiments; the egr3

loss of function mutants was isolated by Igaki et al. (2002).

2.2. RNA analysis

Total RNA from embryos, larvae, adults and S2 cells was extractedusing TRI Reagent® (Sigma) following manufacturer's instruction.RNA was treated with RQ1 DNase (Promega), phenol:chloroformextracted, precipitated with LiCl and resuspended with RNase freewater. For Northern blot analysis, 6 μg of total RNA were electro-phoresed and transferred onto Hybond-NX (Amersham) filters forhybridization. 0.3–0.5 kb genomic DNA fragments spanning theSnoScan detected sequence were used as specific probes; theseprobes were PCR-amplified on genomic DNA by using the appropriateprimer pairs and 32P-labelled using the Nick Translation Kit (Roche).The list of oligonucleotides utilised for Northern blotting or RT-PCRexperiments is shown in Supplemental Table 1.

Basic cloning techniques, PCR amplification, DNA extraction,manipulation and labelling, screening and sequencing techniqueswere carried out according to Sambrook and Russell (2001). The sizeof RNAs was determined by using the High range and Low range RNAMolecular Weight Markers (Fermentas). In quantitative real-time RT-PCR analysis, the RNA (1 μg) was reverse transcribed usingQuantiTect® Rev. Transcription (Qiagen) using manufacturer's con-dition and diluited 1:10. Quantitative real-time RT-PCR experimentswere performed using iQ™5 Multicolor Real-Time PCR DetectionSystem (Biorad). All PCR reactions were carried out in a final volumeof 15 μl using 1 μl of diluited cDNA, 7.5 μl of 2× SYBR® GreenER™(Invitrogen) and 5 pmol of each primer. Sequences of all utilisedprimers were designed using Primer 3 software (http://frodo.wi.mit.edu/) and are shown in Supplemental Table 1. 5.8 S rRNA was usedas endogenous control for samples normalization. Quantitative PCRanalyses were performed using the 2−ΔΔCT method (Livak andSchmittgen, 2001). For 5′-Rapid amplification of cDNA end (5′-RACE), total RNA (2 μg) extracted from S2 cells was reverse-transcribed into first-strand cDNA by using SuperScript III ReverseTranscriptase (Invitrogen) following the protocol supplied. Thesequence of the gene-specific primer (GSP) utilised (P1; see positionin Fig. 5) was: 5′-CCACTGACGTGGGCCTCAA-3′. The removal of RNAcomplementary to the cDNA was performed with 2 U of E. coli RNaseH (Invitrogen) by incubating the sample at 37 °C for 20 min. cDNAwas then purified by using the NucleoSpin Extract II kit (Macherey-Nagel) and eluted in 40 μl of bidistilled water. A homopolymeric A-tail was added to the 3′ end of purified cDNA (10 μl) using 30 U ofrecombinant Terminal Deoxynucleotidyl Transferase (Invitrogen)and 200 μM dATP through incubation at 37 °C for 10 min. The A-tailed cDNA was purified as above and eluted in 20 μl of bidistilledwater. Purified A-tailed cDNA (6 μl) was amplified with an oligo(dT)-adaptor primer (5′-GACCACGCGTATCGATGTCGACTTTTTTTTTTTTTT-TTV-3′) and the GSP 5′-TCAAGTTCGGTTTTCCTCATAAATC-3′ (P2:see position in Fig. 5). A nested PCR was then carried out usingthe nested GSP (P3: 5′-CCGTTAAGCCATGAAGCATT-3′) and theadaptor primer (5′-GACCACGCGTATCGATGTCGAC-3′).

2.3. Cells culture and α-amanitin treatment

S2 cells were grown in Schneider's Drosophila Medium supple-mented with 100 U/ml of penicillin, 100 μg/ml of streptomycin and10% heat inactivated fetal calf serum (Cambrex) and seeded at a

58 G. Tortoriello et al. / Gene 436 (2009) 56–65

4×106/ml concentration. After 24 h, 0.2 μg/ml α-amanitin (Sigma)was added to the culture medium, the cells were harvested 72 h aftertreatment and subjected to RNA extraction.

2.4. TMG immunoprecipitation

Cells were washed in PBS and lysed in buffer A (10 mM HEPES pH7.9,10 mMKCl, 0.1 mM EDTA, 0.1 mM EGTA, 1 mMDTT, 0.5 mMPMSF).After 15 min on ice, NP40 was added to final concentration 0.5%, nucleiwere collected by centrifugation and lysed in RIPA buffer [(150 mMNaCl, 1% NP-40, 50 mM Tris, pH 7.5, 5 mM EDTA, 1 mM PMSF andprotease inhibitor cocktail (Roche)]. Nuclear extracts were incubatedwith theK121 anti-2,2,7-trimethylguanosineMouseAgaroseConjugatemAb (Calbiochem) for 2 h at 4 °C, followed by extensive washes withRIPA buffer. Finally, RNA was extracted from the supernatant andimmunoprecipitated fractions and analysed by Real Time RT-PCR.

2.5. FISH analysis

S2 cells were grown on gelatinized glass coverslips, washed brieflyin phosphate-buffered saline (PBS: 100 mM Na2HPO4, 137 mM NaCl,27 mM KCl pH 7.4) and fixed in PBS containing 4% paraformaldehyde,for 30min on ice. The cells were rinsed twicewith ice-cold PBSM (0.1%MgCl2-PBS buffer) and permeabilized with PBST (0.1%TritonX-100-PBS buffer) for 60 s at room temperature, then rinsed twice withPBSM, equilibrated in pre-hybridization solution (50% formamide/2×SSC) and hybridized for 3 h at 37°C in 20 μl of a mixture containing0.02% BSA, 40 μg of E. coli tRNA, 2× SSC, 50% formamide, 20 ng oflabelled probe. The oligonucleotide probes were: dU85 (Cy3-conjugated), A⁎TAACGTCGTCACCATGACAAACAGCTTAGACCTAACTA;Dm46E3 (Fluorescein-conjugated), G⁎TCTTTCCTCATAAATCTCCGT-TAAGCCATGAAGCAG (Primm). After two rinses with pre-warmedwash solution (50% formamide/2× SSC) at 37 °C for 20min, cells wererinsed first with 2× SSC and then with PBSM at room temperature for10 min. Counter staining of the nuclei was performed by incubatingthe cells with the nuclear staining solution (0.5 μg/ml DAPI in RNasefree water) for 5 min at room temperature. Slides were mounted inMOWIOL 4-88 (Calbiochem).

2.6. Whole mount embryo in situ hybridization

Whole mount embryo in situ hybridization, using single-strandedDIG-labelled probes obtained by PCR, and immunohistochemicalstaining of embryos were performed essentially as describedpreviously (Riccardo et al., 2007).

3. Results

3.1. The list of U79-related Drosophila snoRNAs

Modificationof human28S rRNAat theA3809 residue, guidedby theU79 snoRNA(Smith and Steitz,1998), is highly conserved invertebrates,invertebrates and plants, but absent in yeast. When the SnoScanprogramwasused to search forDrosophilaU79 putative orthologueswedetected, among significant hits, five snoRNAs that were later validatedexperimentally by RT-PCR and Northern analysis (Supplemental Fig.1).The list of detected snoRNAs included – in addition to the three tandemcopies of Me28S-A2634, a snoRNA gene previously molecularlyidentified by Huang et al. (2005) on chromosome arm 3L – two novelsnoRNAs mapping, respectively, on chromosome arms 2R and 3R (seeFig. 1A). According to their position on the polytene map, these novelspecimens were named Dm46E3 (GenBank Acc: EU924798) andDm83E4-5 (GenBank Acc: EU924799), respectively. For each of thesegenes, the antisense element predicted to target A2634, the Drosophila28S residue equivalent to human A3809, was placed upstream theinternal D' box (Fig. 1A).

A case-by-case examination revealed that these genes showedvariegated genomic organizations and expression strategies (Fig. 2).The highly relatedMe28S-A2634 isoforms (named a, b and c) mappedtightly clustered at the 62C polytene region, within a small intergenicregion separating the gene encoding the ribosomal protein RpL23Afrom CG7974, a so far uncharacterized Drosophila gene transcribedfrom the opposite strand. As previously reported, Me28S-A2634 aregenerated by introns of a long, polyadenylated non-coding hosttranscript named dUhg 7 (Huang et al., 2005). This polycistronicarrangement is conserved also in humans where, as mentioned above,U79 is similarly intron-encoded by a long ncRNA, named GAS5, whichharbours 10 box C/D snoRNAs within its introns (Smith and Steitz,1998). Remarkably, structure of a long EST subsequently annotated inthe Drosophila databases (GenBank Acc: BP542227) revealed thatdUhg 7 overlaps with RpL23A and CG7974 genes at their 3′ ends(Fig. 2A), thus showing a more complex arrangement than previouslysuspected. Moreover, we annotatedMe28S-A2634 isoforms also in thecourse of a parallel search for Drosophila U30 orthologues. Sincehuman U30 is responsible for methylation of A3804 (Kiss-Laszlo et al.,1996), a site mapping only 5 nt apart that targeted by U79, we lookedin more detail the Me28S-A2634/28S base pairing properties. Indeed,the complementarity to 28S extended throughout Me28S-A2634 D′boxes exactly by 5 nt, generating a longer duplex of 17 bp. However,this extended tract of complementarity is placed upstream analternative, highly variant D′ box (underlined in Fig. 1B) suggestingthat, instead of guiding site-specific methylation of the A2629 residue,it may support a chaperone activity on rRNA folding. Anyhow,presence of this long antisense element may establish an intermediateevolutive link between the U30 and U79 guiding functions.

The two newly identified snoRNA genes, Dm83E4-5 and Dm46E3,were dispersed in the genome and did not share significativeconservation each other or to the Me28S-A2634 copies. Noticeably,Dm83E4-5 spanned the first exon/intron junction of CG10284(Fig. 2B), a still uncharacterized Drosophila gene predicted to encodea proteinwith serine-type endopeptidase inhibitor activity. In contrastwith the majority of snoRNAs, whose intronic localization permits aconcordant expression with the host gene, its peculiar arrangementpredicts that Dm83E4-5 expression would be alternative to that ofCG10284, being potentially able to negatively regulate host geneactivity. The other snoRNA gene, Dm46E3, was instead embedded,with opposite polarity, within the first intron of eiger (egr; Fig. 2C), afly regulatory gene encoding a Tumor Necrosis Factor-like ligandinvolved in the regulation of programmed cell death, differentiationand immune response (Igaki et al., 2002; Kauppila et al., 2003;Schneider et al., 2007).

When phylogenetic conservation of the five snoRNAs was checkedby multiple genome alignments at the UCSC Genome Browser(http://genome.ucsc.edu), the Me28S-A2634 copies exhibited thehighest sequence conservation, with their presence tracked back tothe D. grimshawi genome. Dm46E3 also displayed significant con-servation within the melanogaster subgroup, whereas Dm83E4-5 wasvery poorly conserved, indicating that it was more recently evoluted.

3.2. Dm46E3 may target a developmentally regulated isoform of thespliceosomal U1 snRNA

Looking carefully at the predicted secondary structures of theduplexes between the Drosophila 28S rRNA and the guide antisenseelements of each snoRNA (see Fig. 1A), we noticed that the duplexformed by Dm46E3 was atypical, being significantly shorter andweakened by two internal mismatches. Both features are expected tolower the efficiency at which the 28S rRNA may be targeted (Chenet al., 2007), prompting us to further investigate whether Dm46E3may play additional functions. Since also spliceosomal RNAs (snRNAs)are known to represent common targets of snoRNA-directed mod-ifications, we checked by BLAST search whether Dm46E3 could

Fig. 1. (A) List of experimentally validated Drosophila U79 orthologues found by SnoScan analysis (Lowe and Eddy, 1999). In the predicted guide duplexes, C/D guide RNA sequencesin a 5′ to 3′ orientation are shown in upper strands (D′ box motifs underlined), while rRNA sequences in a 3′ to 5′ orientation are shown in lower strands. The positions of predictedmethylation sites are indicated by an asterisk. Numbering of Drosophila 28S rRNA is according accession: M21017. OnlyMe28S-A2634 and Dm83E4-5 recognize rRNA target sequencesby complementary antisense elements conforming to the canonical pairing constraints (Chen et al., 2007). (B) Complementarity ofMe28S-A2634 isoforms to 28S extends throughoutcanonical D′ boxes. On the left, alternative base-pairing between Me28S-A2634 isoforms and 28 S rRNA. Canonical D′ boxes, shown in A, are boxed, whereas alternative, highlydegenerate D′ motifs are underlined. The asterisk marks the position of the Drosophila 28S A2629 residue, equivalent to the A3804 site targeted by human U30 (on the right;numbering of human 28 S rRNA is according accession: NR_003287).

59G. Tortoriello et al. / Gene 436 (2009) 56–65

recognize any Drosophila snRNA sequence. Indeed, we observed that itcan target efficiently U1:82Eb (U1b; see Fig. 3A), a developmentallyregulated variant isoform of Drosophila U1 snRNA (Lo and Mount,1990). Intriguingly, the same two nucleotide substitutions thatinterrupt its base pairing to 28S enabled Dm46E3 to match perfectlythe U1b sequence (Fig. 3A), strongly supporting the view that thesetwo point mutations changed the properties of this specimen andcreated a new guiding function. Acquisition of this novel functionalrole would have been facilitated by presence of multiple intact copiesof Me28S-A2634 and Dm83E4-5 genes that preserved the 28S rRNAessential modification function, allowing the Dm46E3 guide elementto evolve more freely and capture U1b as a new target. Among theseven U1 snRNA isoforms documented in D. melanogaster, U1b hasunique structural and expression features. This variant isoform is infact significantly longer than the prototype U1 sequence (designatedas U1a), being characterized by presence of additional 5′ and 3′ tails,

as well as by a single nucleotide change (a G to T transversion atposition 134; bold and underlined in Fig. 3B) within the centralsequence common to all Drosophila U1 isoforms (Lo and Mount, 1990;Fig. 3B). U1b is also characterized by a specific expression profile. Infact, U1a predominates during later stages of the life cycle and invarious terminally differentiated cell lines, whereas U1b is abundantduring the embryonic stages, where it represents the major variantform in cell lines that possess the ability to differentiate (Lo andMount, 1990). Dm46E3 can recognize U1b at its specific 3′ tail, beingpotentially able to modify the terminal region that follows the SMbinding site, at the A234 residue (marked by upper asterisk in Fig. 3B).Worth noting, the Dm46E3/U1b base-paring may elicit functionalconsequences even in the absence of a site-specific modification, forexample by influencing the U1b secondary structure and/or thespliceosomal assembly. To further address the Dm46E3 functionalrole, we then examined whether it localized in the Cajal body (CB), as

Fig. 2. Schematic diagram ofMe28S-A2634, Dm83E4-5 and Dm46E3 genetic arrangement. In A, the structure of the longer EST (Accession: BP542227) representative of the dUhg 7 isoutlined. No EST representative of the Dm83E4-5 and Dm46E3 transcripts was instead found in the databases.

60 G. Tortoriello et al. / Gene 436 (2009) 56–65

predicted for snoRNAs able to guide modifications of the RNA Pol II-transcribed snRNAs, such as U1, U2, U4 and U5 (Richard et al., 2003).As determined by fluorescent in situ hybridization (FISH) of S2 cells,Dm46E3 concentrates within one discrete dot-like nucleoplasmicdomain (Fig. 3C). This region also concentrated dU85 (Richard et al.,2003; Liu et al., 2006), a Drosophila scaRNA known to be enriched inthese nucleoplasmic organelles that we used as CB marker. As shownin Fig. 3D, Dm46E3 sequence also contains Cajal body-specificlocalization signals, or CAB boxes (UGAG; Henras et al., 2004).Altogether, these features indicated that Dm46E3 is indeed a novelDrosophila scaRNA, so hereafter we refer to it as scaDm46E3. Bothlocalisation and base-pairing properties of this novel scaRNA stronglysupport the view that it may effectively be involved in U1b targeting,allowing to predict the first specific modification occurring on thisdevelopmentally regulated U1 isoform.

3.3. Expression strategy of scaDm46E3

Northern blot developmental analysis established that scaDm46E3was constitutively expressed along the Drosophila life cycle (seeSupplemental Fig. 1C). Considering the intriguing antisense arrange-ment of this scaRNA, we attempted to investigate in more detail themolecular mechanisms underlying its expression. No EST with thesame transcriptional orientation was annotated in the databases,suggesting that scaDm46E3 may be independently transcribed as asinglet. Although rare, independent transcription of DrosophilasnoRNAs has been reported to depend on either Pol II or Pol IIImachinery (Isogai et al., 2007), so we wished to address this point bychecking scaDm46E3 sensitivity to α-amanitin. Drosophila cultured S2cells were then treated for 72 h with 0.2 μg/ml of α-amanitin asdescribed under Materials and methods, and the effectiveness of thetreatment determined by real time RT-PCR. In these experiments, thelevel of three Pol II transcripts (αTub and mfl mRNAs, snRNA U1),together with that of the Pol III-transcribed 7SL RNA (Isogai et al.,2007), were followed as internal controls; all data were normalizedagainst the level of the Pol I-dependent 5.8S rRNA, unsensitive to α-aminitin. As shown in Fig. 4B, at the conditions applied the treatmentleft totally unaffected the amount of the 7SL RNA, whereas resulted inan approximately 5-fold reduction of the levels of scaDm46E3 and allPol II-transcribed controls, clearly indicating that scaDm46E3 expres-sion was Pol II-dependent. Given that the majority of snRNAs andsnoRNAs independently synthesized by Pol II has a cap structure

resembling that found on messenger RNAs, but with a characteristictrimethylation of the cap guanosine (Tycowski et al., 2004), we nextchecked whether scaDm46E3 would possess a trimethylguanosine(TMG) cap. Nuclear RNA extracted from S2 cells was then treated withthe monoclonal K121 anti-TMG antibody, and the relative abundanceof scaDm46E3 between the immunoprecipitated or supernatantfractions checked by quantitative real-time RT-PCR analysis. Thelevel of the TMG capped U1 snRNA was followed as positive control,while that of the intron-encoded DmSnR60 snoRNAs, hosted by themfl gene (Riccardo et al., 2007), was taken as a negative reference.As shown in Fig. 3C, the level of U1 was more than 25-fold enriched byprecipitation via the anti-TMG antibody, whereas scaDm46E3 and theintronic DmSnR60 isoforms failed to exhibit any enrichment, pointingout that only a trascurable, if any, fraction of scaDm46E3 shows caphypermethylation. Altogether, these data indicated that scaDm46E3 istranscribed by Pol II most plausibly as a longer, perhaps rapidlyprocessed, precursor. To better define this aspect, we performed 5′RACE experiments to precisely map the scaRNA 5′ end. Noticeably,these experiments revealed that scaDm46E3 transcripts exist as twospecies having distinct 5′ ends mapping more than 400 bp apart (Fig.5A–C). Nucleotide sequencing of the short fragment confirmed the 5′end expected for the scaDm46E3 mature form, whereas that of thelonger, less abundant species revealed presence of an unspliced 5′extension of about 410 nt. Presence of this unspliced extension wasfurther confirmed by RT-PCR analysis, inwhich two upstream forwardprimers (P4/P5) were used in combination with the same reverseprimer (P3) derived from the scaDm46E3 sequence. As shown in Fig.5B, all primer pairs yielded in fact positive amplification of a fragmentwith the expected size. Noticeably, the remote 5′ end mapped incorrespondence of a run of adenines, raising the possibility that theoligo(dT)-adaptor primer may have internally annealed to a longertranscript originating further upstream. Intriguingly, 5′ processing hasrecently been reported for snoRNA:314 and snoRNA:644, twoDrosophila Pol III-transcribed C/D snoRNAs (Isogai et al., 2007).Although the functional significance of these 5′ extensionswas unclear,it is possible that annotation of several snoRNAs is likely to reflect thesize of the mature forms and not the true transcriptional start site.Indeed, specific cleavage of several classes of ncRNAs has recently beenreported, revealing distinct patterns of small RNA biogenesis so farignored (Isogai et al., 2007; Kawaji et al., 2008;Ganesan andRao, 2008)and suggesting that a part of small ncRNAs may be present in multiplelonger forms, which may eventually play different biological roles.

Fig. 3. Dm46E3 target specificity and intranuclear localization. (A) Base pairing of the Dm46E3 D′ antisense element with 28S (up) or snRNA:U:82Eb (U1b: down); the two basesthat interrupt complementarity to 28S (marked in bold) allow perfect matching to U1b. (B) Drosophila U1b sequence; the U1b 5′ and 3′ tails specific tails (capital) flank thecentral prototype U1a sequence (lowercase) containing the SM binding site (underlined) which shows a single G to T transversion (bold capital); the tract of complementary tothe Dm46E3 D′ antisense element (shown in bold capital) contains the A234 target site (marked by the upper asterisk). (C) in situ localization of Dm46E3 RNA in S2 cells.Dm46E3 and dU85 RNAs were visualized by sequence-specific fluorescent oligonucleotide probes, as described in Materials and methods. The merged image shows thatDm46E3 RNA co-localizes with the Cajal body marker dU85 RNA (Richard et al., 2003). (D) Dm46E3 nucleotide sequence. The C box is shown in bold; D and D′ sequencesare boxed; two canonical CAB boxes are underlined.

61G. Tortoriello et al. / Gene 436 (2009) 56–65

Indeed, the observation that scaDm46E3 is embedded within a largeportion of egr intronic sequence that is remarkably conserved acrossthemelanogaster subgroup (Fig. 5D), further supporting the view thatit might derive from a longer precursor.

3.4. Mutual regulation of the scaDm46E3/egr sense–antisense gene pair

Antisense transcription is a widely exploited mechanism toregulate gene expression in eukaryotic cells (Lehner et al., 2002;Osato et al., 2007). Considering that, as demonstrated in many cases,an antisense RNA can inhibit accumulation of another mRNA to whichis complementary, we attempted to investigate in more detail theregulatory mechanisms governing the expression of the scaDm46E3/egr gene pair. In order to assess whether the expression of the twosense–antisense genes could be mutually regulated, we checkedscaDm46E3 accumulation in the egr3 mutant strain, where egrtranscription was shown to be strongly reduced (Igaki et al., 2002).egr3 flies carry a deletion that removes about 1.5 kb of egr genomic

sequences, but leaves the scaDm46E3 coding, upstream and down-stream surrounding sequences intact (Igaki et al., 2002; see alsoFig. 6A). When the scaDm46E3 expression level of wild-type and egr3

flies was compared by quantitative real-time RT-PCR analysis, wefound that it wasmore than 10-fold increased in themutants (Fig. 6B),indicating that egr transcriptional efficiency can significantly influencethe expression of its antisense snoRNA. Moreover, the finding thatscaDm46E3 accumulates at an even higher amount in S2 culturedcells (Fig. 6B) suggested that its expression may be up-regulated inproliferating cells, or eventually down-regulated in adult flies byregulatory factors not present in this cell line. Mutual regulation bysense–antisense transcription may make gene expression amenablenot only to temporal, but also to spatial regulation, so that RNAs thatform sense–antisense pairs frequently exhibit reciprocal expressionpatterns. In the case in point, egr is known to exhibit a specificexpression pattern during embryogenesis and in larval tissues, beingpredominantly expressed in the nervous system (Igaki et al., 2002;Kauppila et al., 2003). We thus asked whether scaDm46E3 might

Fig. 4. ScaDm46E3 is Pol-II-dependent and lacks Trimethylguanosine cap. (A) Sensitivity of scaDm46E3 accumulation toα-aminitin, showing its dependence on Pol II machinery. RNAfrom untreated or treated (see Materials and methods) cells was analysed by Real time RT-PCR. Data were normalized to 5.8 rRNA expression and presented relative to the untreatedcells; they represent the average of at least three independent experiments. In treated cells, scaDm46E3 expression decreases similarly to the examined Pol II-dependent transcripts(αtub, mfl, snRNA U1), whereas the Pol III-transcribed 7SL RNAwas totally unaffected. (B) Quantitative real-time RT-PCR analysis on the immunoprecipitated (IP) or supernatant (S)fractions of nuclear RNA extracted from S2 cells after incubationwith the K121monoclonal anti-TMG antibody (Calbiochem). Trimethylguanosine capped U1 snRNA shows a markedIP-enrichment, whereas no IP-enrichment was observed for scaDm46E3 or the DmsnR60 negative control.

62 G. Tortoriello et al. / Gene 436 (2009) 56–65

similarly display a spatially-restricted expression profile duringembryogenesis. To this purpose, staged whole-mount embryopreparations were analysed by in situ hybridization withscaDm46E3-specific probes. As shown in Fig. 7A, a strong ubiquitoushybridization signal was detected in pre-blastoderm embryos (stages1–3), indicating a high level of maternal contribution. Consideringthat U1b is maternally deposited in the developing oocytes (Lo andMount, 1990), the high-level expression of scaDm46E3 observed inearly embryos nicely correlates with its predicted role on U1bmodification. Immediately after, scaDm46E3 expression concentratedat posterior pole, although it remained excluded from the pole cells,the first cells formed in the Drosophila embryo and the germ cellprecursors (stage 4; Fig. 7B). During early gastrulation (stages 6–7;Fig. 7C) the signal marked the amnioproctodeal invagination, wherethe primordia of the hindgut and the posteriormidgut are formed, andremained concentrated there also in the following stages. High-levelexpression of scaDm46E3marked in fact these tracts of the developingintestine also during late gastrulation and in segmented embryos(stages 9–10; Fig. 7D, E). At these stages, scaDm46E3 also marks theneurogenic region, a territory in which its expression coincides with

that of egr (Igaki et al., 2002; Kauppila et al., 2003). Finally, in latesegmented embryos scaDm46E3 is also prominently upregulated inthe amnioserosa (stage 14; Fig. 7F), a dorsal stretch of extraembryonicepithelium that connects the dorsal and ventral halves of thesegmented germ band and is progressively covered by lateralectodermal cells undergoing dorsal closure, the last morphogeneticmovement during embryogenesis. An obvious conclusion from thesein situ experiments is that the spatial expression of scaDm46E3 isdynamically regulated during embryogenesis. Moreover, it concen-trates in the developing intestine and in the amnioserosa, twoembryonic regions not marked by egr, whose expression is restrictedat the dorsal furrows until embryonic stage 9, and thereafterconcentrates essentially in the developing nervous system (Igakiet al., 2002; Kauppila et al., 2003).

4. Discussion

Although ribose 2′-O-methylation represents the most commonnucleotide modification of rRNAs and snRNAs, a significant portionof C/D snoRNA genes still remains to be identified in most genomes.

Fig. 5. ScaDm46E3 transcripts exist as two different species differing at 5′ end. (A) 5′ RACE products obtained after nested PCR-amplification (performed with the P3/adaptor primerpair; see Materials and methods). For 5′ RACE, retrotranscription makes use of the P1 gene-specific primer, and the P2/oligo(dT)-adaptor primer pair was used in the first PCRamplification (see (C) for position of primers). A 100 bp amplification product corresponding to the scaDm46E3 mature form and a highest accessory band are detected in presence(+RT), but not in absence of reverse transcriptase (-RT), indicating absence of contaminating genomic DNA. Lane M: 100 bp Plus Molecular Weight Marker (Applichem). (B) RT-PCRanalysis on S2 RNA confirming the presence of scaDm46E3 long precursor form. Both P3/P4 and P3/P5 primer pairs show positive amplification of fragments with the expected sizein presence (+), but not in absence of reverse transcriptase (−), indicating absence of contaminating genomic DNA. (C) Schematic representation of the primers used in A and B. (D)Blat search of egr exon1–intron1–exon2 sequence indicates a marked conservation of a large intron 1 portion across the melanogaster subgroup.

63G. Tortoriello et al. / Gene 436 (2009) 56–65

Indeed, identification of specimens showing atypical structure orgenomic location, poor expression or unknown function is expectedto be particularly elusive. Our results testify how a combination ofbioinformatics, comparative genomics and molecular approachesmay synergistically contribute to undercover the hidden layer of thesnoRNA transcriptome, extending our knowledge about snoRNAgenes, their genomic arrangement, expression strategies andevolutive mechanisms. In previous works, these combinedapproaches already proved to be successful, allowing us to identifythe Drosophila counterparts of two out of ten GAS5 intron-encodedsnoRNAs (Accardo et al., 2004; Riccardo et al., 2007). In addition toMe28S-A2634 isoforms, which may represent ancient U79 ortholo-gues, two novel RNAs, snoDm83E4-5 and scaDm46E3, were identifiedin the course of our bioinformatics search. Duplication and dispersionin the genome, followed by genetic drift, or convergent evolution arepossible mechanisms accounting for the origin of these dispersedgenes. An additional mechanism has been recently evoked for theevolution of mammalian snoRNAs, for which retroposition of aparental copy, followed by genetic diversification, has been hypothe-sized (Weber, 2006; Luo and Li, 2007). Whatever process creates anew snoRNA copy and establishes genetic redundance, over evolu-tionary time divergence dynamic mechanisms can increase snoRNAdiversity to eventually give rise to an inactive isoform that woulddecay during the course of evolution or, alternatively, originate a newRNA-modification function that might be lineage or species specific.

To this regard, it is widely accepted that point mutations in motifsrelated to C, D or D′ box sequences or in the antisense elementsmight have a strong impact on the functionality of a snoRNA copy,playing a key role on the evolution of novel functions. By thismechanism, snoRNAs may evolve new target site complementaritiesand the gain, loss and change of targets of over relatively shortevolutionary times may allow these small ncRNAs to continuouslycontribute to the changing needs of cells and genomes. In this light,the scaDm46E3 gene described nicely shows how subtle changeswithin the antisense element can drastically affect the target speci-ficity, creating a new guide function. Two point mutations within thescaDm46E3 antisense element were in fact sufficient to produce atarget switch from 28S rRNA to the U1b snRNA. Although a moreconclusive verification of U1b modification is necessary, it can besurmised that even in the absence of site-specific modification thescaDm46E3/U1b base-paring may elicit functional consequences,possibly by influencing snRNP assembly. Intriguingly, out of the 63fly C/D box snoRNAs identified so far, only five were expected toguide internal methylations of snRNAs, with U2, U5, and U6 snRNAsrepresenting the predicted targets (Huang et al., 2005). Thus, thescaDm46E3 antisense element leads to envisaging the first methyla-tion on Drosophila U1 snRNAs, with the U1b embryonic variantexpected to be specifically targeted. Intriguingly, the finding thatscaDm46E3 is expressed with a dynamic, spatially restricted expres-sion profile during Drosophila embryogenesis raises the possibility

Fig. 6. ScaDm46E3 is over expressed in egr3 mutant flies and in S2 cells. (A) Schematic diagram of the 1.5 kb deletion carried by egr3 flies and relative position of the scaDm46E3 gene.(B) Amounts of scaDm46E3 transcripts inwt and egr3 adult flies and S2 cells, determined by real-time RT-PCR. Data represent the average of at least three independent experiments.The markedly higher scaDm46E3 accumulation observed in the egr3 strain indicates that egr transcriptional efficiency influences the expression of its antisense scaRNA. The evenmore pronounced scaDm46E3 accumulation level observed in S2 cell suggests that this snoRNA is up-regulated in proliferating cells, or down regulated in adult flies.

64 G. Tortoriello et al. / Gene 436 (2009) 56–65

that a subset of snRNA modifications might be developmentallyregulated. According to this hypothesis, it seems conceivable thatsnoRNAs able to target spliceosomal RNAs may act as critical splicingmodifiers, mediating in a subtle way a variety of biological effects inhigher eukaryotes.

In the evolution of a snoRNA copy, the activity of a new gene oftenrelies on the establishment of a novel expression strategy that, onceestablished, may exert diverse regulatory effects on overlapping orsurrounding genetic regions. Intriguingly, both the snoDm83E4-5 andscaDm46E3 genes show unusual genomic location and followexpression strategies totally different from that displayed by themajority of Drosophila C/D snoRNAs. While most C/D snoRNAs aregenerally intron-encoded by either protein-coding or non-coding

Fig. 7. ScaDm46E3 shows a dynamic expression profile during Drosophila embryogenesis. Drdevelopmental stages. In (A) and (B) pre-blastoderm embros; the strong ubiquitous signsegmented embryos (D, E) and late embryos undergoing dorsal closure (F). The signal conamnioserosa, before these cells are covered by the dorsal closure.

host genes (HGs), and released after splicing of HG's primarytranscript, expression of snoDm83E4-5 and scaDm46E3 is insteadenvisaged to play a negative role on the activity of their overlappedgenes. Based on its location at an exon–intron junction, snoDm83E4-5expression is in fact predicted not only to be splicing-independent,but even alternative to the production of its HG's spliced mRNA.Similarly, expression of the scaDm46E3/egr sense–antisense genepair appears to be mutually regulated, making plausible thatscaDm46E3 expression may regulate egr activity via RNAi or bymodulating splicing of egr pre-mRNA. A significant percentage ofseveral genomes, including those of mammals, is known to betranscribed from both strands, and a number of well-characterizedantisense transcripts appear to play regulatory roles in relation to

osophila embryos were labelled with digoxigenin-labelled antisense probes at differental in (A) denotes maternal contribution. (C–F) lateral views of early gastrulating (C),centrates from the primordia in the hindgut and the posterior midgut, and later in the

65G. Tortoriello et al. / Gene 436 (2009) 56–65

their sense gene (reviewed by Lapidot and Pilpel, 2006). In additionto work by RNA-directed mechanism, symmetrical transcription maybe critical in regulating the state of a large chromosomal domainaround a gene, leading to alterations of chromatin structure, DNAmethylation, as well as promoter exclusion or competition. Althoughfurther experiments are required to understandingwhatmechanism isactually responsible forDm46E3/egrmutual regulation, it reasonable toassume that it plays a role in ensuring that these two genes maintainthe distinct spatial expression profiles observed during embryogenesis.Considering that, as member of the tumor necrosis factor (TNF) family,egr plays an important role in the regulation of cellular proliferation,differentiation and apoptosis through the activation of the JNKpathway(Igaki et al., 2002), further characterization of scaDm46E3 may berelevant to understanding the molecular mechanisms governing theexpression of this fly regulatory gene. Indeed, several DrosophilasnmRNAs displayed active expression when located in an antisenseorientations (Yuan et al., 2003; Accardo et al., 2004; Huang et al., 2005)suggesting that the small ncRNA transcriptome might have broadimplications for fly gene regulation.

Acknowledgments

We thank M. Dionne for furnishing the egr3 strain and for helpfuldiscussion. This work was supported by funds from the Assessorato allaRicerca Scientifica, Regione Campania (Legge 5, annualità 2006) to M.F.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at doi:10.1016/j.gene.2009.02.005.

References

Accardo, M.C., et al., 2004. A computational search for box C/D snoRNA genes in theDrosophila melanogaster genome. Bioinformatics 20, 3293–3301.

Bachellerie, J.P., Cavaille, J., Huttenhofer, A., 2002. The expanding snoRNA world.Biochimie 84, 775–790.

Chen, C.L., Perasso, R., Qu, L.H., Amar, L., 2007. Exploration of pairing constraints identifies a9 base-pair core within box C/D snoRNA–rRNA duplexes. J. Mol. Biol. 369, 771–783.

Dong, X.Y., et al., 2008. SnoRNA U50 is a candidate tumor-suppressor gene at 6q14.3with a mutation associated with clinically significant prostate cancer. Hum. Mol.Genet. 17, 1031–1042.

Ganesan, G., Rao, S.M., 2008. A novel noncoding RNA processed by Drosha is restrictedto nucleus in mouse. RNA 14, 1399–1410.

Henras, A.K., Dez, C., Henry, Y., 2004. RNA structure and function in C/D and H/ACA s(no)RNPs. Curr. Opin. Struct. Biol. 14, 335–343.

Huang, Z.P., Zhou, H., He, H.L., Chen, C.L., Liang, D., Qu, L.H., 2005. Genome-wide analysesof two families of snoRNA genes from Drosophila melanogaster, demonstrating theextensive utilization of introns for coding of snoRNAs. RNA 11, 1303–1316.

Huang, Z.P., Chen, C.J., Zhou, H., Li, B.B., Qu, L.H., 2007. A combined computational andexperimental analysis of two families of snoRNA genes from Caenorhabditis elegans,revealing the expression and evolution pattern of snoRNAs in nematodes.Genomics 89, 490–501.

Igaki, T., et al., 2002. Eiger, a TNF superfamily ligand that triggers the Drosophila JNKpathway. EMBO J. 21, 3009–3018.

Isogai, Y., Takada, S., Tjian, R., Keles, S., 2007. Novel TRF1/BRF target genes revealed bygenome-wide analysis of Drosophila Pol III transcription. EMBO J. 26, 79–89.

Kauppila, S., et al., 2003. Eiger and its receptor, Wengen, comprise a TNF-like system inDrosophila. Oncogene 22, 4860–4867.

Kawaji, H., et al., 2008. Hidden layers of human small RNAs. BMC Genomics. 9, 157.Kishore, S., Stamm, S., 2006. The snoRNA HBII-52 regulates alternative splicing of the

serotonin receptor 2C. Science 311, 230–232.Kiss, T., 2002. Small nucleolar RNAs: an abundant group of noncoding RNAs with

diverse cellular functions. Cell 109, 145–148.Kiss, T., Fayet, E., Jady, B.E., Richard, P., Weber, M., 2006. Biogenesis and intranuclear

trafficking of human box C/D and H/ACA RNPs. Cold Spring Harb. Symp. Quant.Biol. 71, 407–417.

Kiss-Laszlo, Z., Henry, Y., Bachellerie, J.P., Caizergues-Ferrer, M., Kiss, T., 1996. Site-specific ribose methylation of preribosomal RNA: a novel function for smallnucleolar RNAs. Cell 85, 1077–1088.

Lapidot, M., Pilpel, Y., 2006. Genome-wide natural antisense transcription: coupling itsregulation to its different regulatory mechanisms. EMBO Rep. 7, 1216–1222.

Lehner, B., Williams, G., Campbell, R.D., Sanderson, C.M., 2002. Antisense transcripts inthe human genome. Trends Genet. 18, 63–65.

Liu, J.L., Murphy, C., Buszczak, M., Clatterbuck, S., Goodman, R., Gall, J.G., 2006. TheDrosophila melanogaster Cajal body. J. Cell Biol. 172, 875–884.

Livak, K.J., Schmittgen, T.D., 2001. Analysis of relative gene expression data usingreal-time quantitative PCR and the 2(−Delta Delta C(T)) method. Methods 25,402–408.

Lo, P.C., Mount, S.M., 1990. Drosophila melanogaster genes for U1 snRNA variants andtheir expression during development. Nucleic Acids Res. 18, 6971–6979.

Lowe, T.M., Eddy, S.R., 1999. A computational screen for methylation guide snoRNAs inyeast. Science 283, 1168–1171.

Luo, Y., Li, S., 2007. Genome-wide analyses of retrogenes derived from the human boxH/ACA snoRNAs. Nucleic Acids Res. 35, 559–571.

Mattick, J.S., 2004. The hidden genetic program of complex organisms. Sci. Am. 291,60–67.

Mattick, J.S., Makunin, I.V., 2006. Non-coding RNA. Hum. Mol. Genet. R17–R29 15 Spec.No. 1.

Mourtada-Maarabouni, M., Hedge, V.L., Kirkham, L., Farzaneh, F., Williams, G.T., 2008.Growth arrest in human T-cells is controlled by the non-coding RNA growth-arrest-specific transcript 5 (GAS5). J. Cell. Sci. 121, 939–946.

Mourtada-Maarabouni, M., Pickard, M.R., Hedge, V.L., Farzaneh, F., Williams, G.T., 2009.GAS5, a non-protein-coding RNA, controls apoptosis and is downregulated in breastcancer. Oncogene 28, 195–208.

Nakamura, Y., et al., 2008. The GAS5 (growth arrest-specific transcript 5) gene fuses toBCL6 as a result of t(1;3)(q25;q27) in a patient with B-cell lymphoma. CancerGenet. Cytogenet. 182, 144–149.

Osato, N., Suzuki, Y., Ikeo, K., Gojobori, T., 2007. Transcriptional interferences in cisnatural antisense transcripts of humans and mice. Genetics 176, 1299–1306.

Pontes, O., Pikaard, C.S., 2008. siRNA and miRNA processing: new functions for Cajalbodies. Curr. Opin. Genet. Dev. 18, 197–203.

Riccardo, S., Tortoriello, G., Giordano, E., Turano, M., Furia, M., 2007. The coding/non-coding overlapping architecture of the gene encoding the Drosophila pseudouridinesynthase. BMC Mol. Biol. 8, 15.

Richard, P., Darzacq, X., Bertrand, E., Jady, B.E., Verheggen, C., Kiss, T., 2003. A commonsequence motif determines the Cajal body-specific localization of box H/ACAscaRNAs. EMBO J. 22, 4283–4293.

Rogelj, B., 2006. Brain-specific small nucleolar RNAs. J. Mol. Neurosci. 28, 103–109.Sambrook, J., Russell, D.W., 2001. Molecular Clonning: A LaboratoryManual. Cold Spring

Harbor Laboratory Press.Schneider, D.S., et al., 2007. Drosophila eiger mutants are sensitive to extracellular

pathogens. PLoS Pathog. 3, e41.Smith, C.M., Steitz, J.A., 1998. Classification of gas5 as a multi-small-nucleolar-RNA

(snoRNA) host gene and a member of the 5′-terminal oligopyrimidine genefamily reveals common features of snoRNA host genes. Mol. Cell. Biol. 18,6897–6909.

Stanek, D., et al., 2008. Spliceosomal small nuclear ribonucleoprotein particlesrepeatedly cycle through Cajal bodies. Mol. Biol. Cell 19, 2534–2543.

Tycowski, K.T., Aab, A., Steitz, J.A., 2004. Guide RNAs with 5′ caps and novel box C/DsnoRNA-like domains for modification of snRNAs in metazoa. Curr. Biol. 14,1985–1995.

Vitali, P., et al., 2005. ADAR2-mediated editing of RNA substrates in the nucleolus isinhibited by C/D small nucleolar RNAs. J. Cell Biol. 169, 745–753.

Weber, M.J., 2006. Mammalian small nucleolar RNAs are mobile genetic elements. PLoSGenet. 2, e205.

Yuan, G., Klambt, C., Bachellerie, J.P., Brosius, J., Huttenhofer, A., 2003. RNomics inDrosophila melanogaster: identification of 66 candidates for novel non-messengerRNAs. Nucleic Acids Res. 31, 2495–2507.